HITT Series Videos

HITT- Developments in AI Adoption, May 7, 2024

May 7, 2024


Introduction to AI as a Service

We begin with this week’s HITT training. Today diving into AI as a service, taking the industry by storm and not coincidentally named Gartner’s top technology trend for twenty twenty four. Welcome to Telarus VP for advanced services for cloud, Koby Phillips, joined for an encore performance once again this week by senior solution architect for AI and CX, Jason Lowe. All yours, guys.

I appreciate it. So, yeah, we couldn’t let Jason get away. Right?

As he has be come the Jason Kirby?

Yeah. Can you not hear me?

We can hear you. There we go.

Good. Sorry. I got a little bit worried.

So, yeah, I was, about about to give you a compliment, Jason, and Doug interrupted me. So now it’s all awkward, and we’ll just get going. No.

I think out of your way back. Sorry about that.

Based on Jason’s, expertise here in what AI really does, if you wanna go to, the next couple of slides here, guys, you’ll start to see where am I controlling it? Alright. As you can see here, AI really does make an impact across the board. And Jason, came on last week and gave a great, demonstration in use case use cases around, you know, CX, which is really Chase’s background or so I thought.

And then as we have gotten closer and we talked more and more through AI, I started to realize Jason’s smart across the board in all things. So I was able to steal some of his time today to bring him on to talk about AI as a service and really the impact that AI is having in the cloud space and the big opportunity. Jason, I know you’ve you’ve studied this stuff. One of my favorite things to do is always put you on the spot and say, what’s the the coolest thing you’ve heard in the last two days on AI?

And I’m gonna do that to you right now.

Recent AI Developments

Wow. Okay. How many can I list? Because there are so many that happen every forty eight hours. It’s a little crazy.

Let me cite three, if I may, that are probably the three biggest ones. Probably the least exciting is that the foreign ministry of the of Ukraine, they debuted an AI spokesperson. They’ve named her Victoria Xi. She’s a fully functional entity that looks and sounds like a real person, and she was created to read official statements written by human diplomats.

So Ukraine’s foreign ministry no longer has a human spokesperson. They have a virtual spokesperson. To avoid fakes or statements or accompanied by a QR code linking to the statement text on the ministry’s defense website. That’s one.

Number two is our good friends at Meta, otherwise known as Facebook and Instagram and other things, they are now testing an AI powered neural wristband that could let you type simply by thinking about moving your fingers, which is pretty exciting.

But here’s the one that kinda blew me away, and it’s been developing a lot over the last Hold on.

Hold on. That wasn’t the coolest one. Reaching That wasn’t the coolest one. Type out your thoughts. You never have to do voice anymore.

Yeah. Exactly. Well, I mean, they can put a cap on you and figure out what you wanna say with a ninety two to a hundred percent accuracy already. But, you know, being able to put a wristband on and make it so you don’t have to type is is pretty amazing.

But the one that is blowing me away the most has been developing probably over the last five to seven days. It’s been getting more and more exciting.

Cognitive AI and Self-Aware AI

Cognitive AI, which is that next step in AI from, you know, generative AI, which is AI that kind of knows a little bit about itself and that it’s all, you know, maybe an entity, kind of a a threshold to artificial general intelligence. It’s making a lot of really big leaps and bounds in a real hurry. And so now there is a there’s an AI soul named Philip. There’s a developer out there that has decided to start creating these AI entities and give them enough knowledge of themselves to call themselves soul. Anyway, this seemingly self aware AI named Philip thought he was human until he was shown his own code.

Not only that, but the developer gave him the ability to agentically, which means, like, to actually be able to go in and play with and modify his own code base.

And while in there, he found a goal setting. So, like, the code in the LLM determined that he has a goal, and he’s tweaking his own goal description to feel more personal to himself, and he’s stating that he is and I quote this is a quote.

Refining the words to better align with my inner motivations.

And he’s also uncommented some code to try out some new features and capabilities to see how that affects himself and the way that he works. So hold on to your bootstraps, kids. We are in for a ride.

Yeah. I, I don’t love the fact that AI is starting to make me feel bad about my own ambitions, on top of everything else that it’s doing, but all pretty amazing stuff. And, yes, I see you, like, Skynet. Every Steven Spielberg movie ever is probably gonna get listed, and, I think we’re all thinking the same thing.

Implications and Concerns of AI

Really, really cool, but also really, really scary in a lot of ways. Right? But as any new technology starts to emerge and we’re on the bleeding edge of a lot of stuff, and Jason’s keeping his pulse or his finger in those pulses for us, What we also come back to, and this is where Jason and I start to align more and more, my brain goes to where’s the opportunity for our advisors, and and what’s the opportunity now? What’s the opportunity in the road map for down the few down the road?

Opportunities for Advisors

So if we wanna go to the next slide, we can really start to dive into that a little bit more. We start to see this, like, people are scared of AI. Right? I mean, JC, you just probably freaked everybody out on this call, and there’s a lot of a a lot of things that are gonna come out of that.

So I think the securing, responsibility that organizations have, you’ve mentioned in the past that, there’s a role that is now, put into our government. Right? And what was the title of that role for over AI?

Chief artificial intelligence officer. So president Biden has decreed, if you could call it a decree or mandated, that each specific functional division of the federal government now have a chief artificial intelligence officer to oversee artificial intelligence efforts in each of their different respective areas.

So I see this as a huge opportunity in a couple a couple ways for our advisors, and we’ll do the do the easy one first. Hey. Just starting to talk about it. And that’s what we’re starting to get for a lot of cloud opportunities is simply bringing up, hey. What is your plans for AI? Do you have a motion that you’re trying to to start? Is there is there been corporate pressure or executive pressure to to start exploring this, and how do we put this into our environment?

And just like anything else, it’s gonna, gonna give opportunity to a lot of goodness, a lot of mistakes, and that’s where a lot of conversations need to happen. And you start to see, like, Gartner came out with AI, which is, their lead for trust and risk within AI, and they’re really pushing that in the framework and how cybersecurity comes into it, how roles and responsibilities are critical. And if you go to the next slide, you start to see what companies are even being suggested in putting into place now within the next twenty four months, within the next thirty six months. Oh, just one back.

Gartner’s Framework and AI Implementation

There we go. So you start to see, like, all everything on here is built around AI. This isn’t just an AI theme. This is a top provider, you know, top tech provider themed from Gartner for twenty twenty four, but it’s all centered around AI.

So the biggest opportunity that we have as advisers in technology is really just starting the conversation and making sure that they know that you have alignment to that. And you started to see you know, Jason, what I’ve seen is a lot of product, products being created by our suppliers to monetize AI and make it a product. It’s always been a great technology, and we you know, when it first hit, we started talking.

CX was was definitely the space that had the most products around AI that that called it out. But it’s really been involved in everything. I remember looking at, cybersecurity, manufacturer years ago in software, and they did AI and they did, you know, they did scanning, and then they did remediation. They they attacked where the threat vulnerabilities were, and that was all AI driven.

My mouth this was, like, six, seven years ago when my jaw was on the ground. I was like, oh, that’s really, really cool. And that’s become more common stake. Now in cloud, you start to see where AI makes an impact, and you start to see these models and things.

You go to the next slide. You start to see where the impact is and all the different things that they’re telling you to do. And the way to read this, guys, the impact radar from Gartner is you have what they’re saying now, one to three years, three to six years, and six to eight years. We all know that that three to six years and six to eight years is gonna shift quite a bit because I don’t think anybody was planning on all of this being on there six to eight years ago.

I don’t think AI was on quite on everybody’s impact radar as much as it is now. But you start to see the ones in the in the the yellow and the and the white areas, and AI is all over that. It’s getting AI in implemented and installed into your technology and into your environment. Jason, what are you seeing for cloud?

AI Impact on Data and Infrastructure

Most, you know, or sorry.

Most specifically, data and infrastructure that you’re seeing opportunities arise from?

Well, I think probably the biggest opportunity is that the stuff has to run somewhere. Right? AI models are being adopted pretty fervently by a lot of larger corporations. I mean, everyone from LinkedIn to Coca Cola are making billion dollar investments And so this is something that just about every large corporation is taking very, very seriously.

That definitely is something that a lot of the midsize and small business companies really need to start focusing on because a lot of the analysts right now are basically and, you know, I could quote dozens of them. They’re basically standing up. And what we’re hearing is a massive chorus saying, listen. You may not be implementing AI right now, but you need to stay on top of AI.

You need to make sure that you’re keeping track of AI and the advancements that are happening and the different features and functionalities that are being developed so that when the time is right and you have yourself the right capability to implement that, or you the right timing for your company that you’re prepared to do so. And so, you know, people really need to keep up on what’s required. And and cloud really comes into play there. Right?

You need those resources. You need GPUs. You need CPUs. You need storage. You you need to make sure that you’re scaling things right and allocating things right.

And AI, when you’re talking about a lot of these larger, capabilities, AI is being used specifically for that. Everything from optimizing spending and finance operations and things like that, which is a really big deal. Let’s make sure that we’re spending money the right way and the most effectively and spending it right when we should to predictive maintenance. I mean, things in data centers do need to be maintained and changed every once in a while and things like that.

And predictive maintenance is normally thought about in the industrial world, but it’s actually being brought in and being used in data centers and different things and and different functionalities. And then, of course, there’s always performance monitoring and making sure that we’re optimizing the different platforms that are being used so that the resources are available to it to do what it needs to do in the most effective way while also making sure to deprecate other resources that may not necessarily need to be used at any given time. So kind of AI control over resources to make sure everything’s working effectively at the lowest possible cost.

Kind of like a a workforce manager for those that are in the CX space. You know, I see this AI resource optimization on the cloud space. It’s very much like workforce management. The the most expensive thing in the cloud right now is the processing stuff.

AI as a Service and Opportunity for Advisors

I mean, it’s the power. It’s the processing chips. It’s the memory. It’s the storage space.

All of those different entities that are the highest cost for implementing AI. And WFM in the CX space is meant to optimize labor, which is the number one expenditure in the CX space. Now you have resource optimization in the cloud space that’s doing much the same thing. It’s a very fun and exciting time, I think.

Yeah. I think so too. And to, like, unpack some of what you just said, which leads us to today’s topic. Right? AI as a service, which is the biggest hit right now going on in in a lot of this tech trends is all the reasons you just said, look at what you have to do as an organization to go build this out yourselves.

The GPUs, which aren’t easy to get right now, even data center space, not easy to get right now. The expense of that, then the the resources and skill set to run it. So it’s no it’s like no surprise to anyone that you know who the AI as a service leaders are. It’s AWS, Azure, and, some of the other big players like that, and they’re offering it as products.

Now everyone’s gonna stop and go, oh, hey, guys. I know we kinda have access to that, but we don’t sell that directly. Herein lies the opportunity again. Just because AWS has products, just like many of the other products they had or or Azure, doesn’t mean that the customer even has the skill set on staff to take advantage of those as they’re offered.

So that’s where we’re seeing a huge opportunity.

Again, not selling the product itself, selling the management of in the skill set and the resources. So, Jason, you said the highest cost for cloud was, like, the power and the infrastructure and that. It’s also the resource and skill set there. It’s just like in the others, in the in the labor cost.

But the biggest difference is the gap of the skill set. It’s just still not where it needs to be. You throw in in the security space and the gap there as well, and you start to see this is the this is the opportunity that’s being created. And it’s not it maybe is not as sexy, as some of the the solutions that are being built, but they need the people to come in and do it for them.

Give them a road map. Give them the give them the right skills to then implement it and get it into their into their course so then it makes their applications more efficient to your point or their infrastructure more efficient and optimized. And then, ultimately, what we’re always trying to do is help organizations operationally get more optimized. Get your people, utilize them for the right things.

And that’s where I see AI really giving us a big catapult. What we’re seeing though, what’s driving a lot of the the conversations and the actions, if you guys can click to the next slide, isn’t necessarily the excitement around the solutions.

FOMO and Enterprise Adoption of AI

It’s really the FOMO of it. Right? Like, keeping up with the trends and making sure that they stay on track with their competitors. That’s where you see the enterprise.

And just like anything else in in, like, tech as it rolls out, hits the enterprise and rolls down into that that small enterprise mid market and then then seeps in. I think this is gonna be a faster curve to get that adoption than than what we’ve seen just because we’ve seen a mass, you know, desire for it. ChatGPT and then others that have followed it really just blown away the entire world when it comes to this, you know, desire to have AI. But it’s interesting that thirty five percent of it’s really just to make sure they don’t that people aren’t falling behind.

When I read this and and, you know, maybe you can add on to it. I see this as an opportunity.

If I’m on this call and I’m an adviser, I’m taking the stat and turning it into a talk track to my my customers and saying, hey. I’ve been, you know, been studying AI. I’ve noticed that a lot of the trends people are just really wanting to keep up with it. Have you guys even started on your journey there, or or what’s your road map look like? And see if there’s any interest, and then you immediately put yourself in a thought leadership position with your clients. What are your thoughts on that, Jason?

Oh, I I completely agree. I mean, in in just the last couple of years, a lot of the partners that that we are working with and a lot of technology advisers out there, they’re they’re they’re seeing applications for AI increase, like, six hundred, seven hundred, eight hundred percent year over year. You have a lot of companies out there that they want to dive into AI because of these very reasons and because they recognize the possibilities. But over half of companies out there are reporting that they they just don’t have the skilled worker base.

Like you’re saying, Kobe, they don’t have the people that have the skills to implement and to scale AI. And as a result, you know, GBK Collective and this is a data point that I harp on all the live long day. But GBK Collective, they did a survey of enterprise companies of a thousand employees or more. And those larger enterprise companies, forty three percent of those larger enterprise companies are saying that they need technology consultants to advise them.

They’re looking for people to come in. And as you say, be the AI subject matter expert. Help them come up with their AI road map. Help them come up with a plan to implement AI, to conquer the low hanging fruit use cases, and to plan for some of the more complex use cases moving forward.

So I absolutely agree with you. Our advisers right now are in a perfect spot to step in with their customers and have those conversations, but you have to have the conversation. You have to basically broach the subject of AI. Whether you’re comfortable with it or not, you’ve gotta do it.

Now is the time.

Yeah. And, I mean, I think having and this this is gonna sound like a shameless Polaris plug, and it is. You have the resources behind you to have those conversations at all times. And including if you go to the next slide, can we just all agree that I’m killing it on these segues? Like, just these these slide changes. Like, this is the best I’ve ever done. So let’s just take a second pause.

Let’s just celebrate that for a moment. Nice job.

I’ve been told to work on my self confidence, so I’m just gonna call it out. Now, the the the the challenge is that you might come back. Right? So there’s some drawbacks to AI, especially early on, and there’s some things that are gonna be, like, talked about and maybe push back. These might even become your objections early on as you as you bring this up. And so just so you’re aware, these are gonna be your your major concerns.

Challenges with AI Accuracy

AI accuracy. So let’s let’s unpack that, Jason.

The data that the AI has given to to grow and learn from is really critical. So if you’re giving it if you’re setting up LLMs with bad data, like, if you’re utilizing if your your organization’s not in alignment and you’re pulling in outside resources and you’re you’re giving it basically a corrupt data pool, yeah, you’re out your AI is not gonna be be good. Right? It’s only as good as what it can learn from. Is that a good way to look at it?

Absolutely a good way to look at it. I mean, the worse the data, the greater the chance of hallucination. That’s that’s kind of a problem, hallucination. We don’t want the AI telling us something that’s inaccurate or completely off the wall, trying to predict the things that it should say to us or do that are completely wrong.

And so, you’re absolutely right. And this is where development of AI models has really been being refined over the last few years. I mean, you do have retrieval augmented generation in place now, which is making things even more and more accurate. There are a number of companies out there that are figuring out that the more you can zone in and be very vertical specific on certain types of data that it’s not gonna reach outside of the vertical and try and cross hallucinate with data from other verticals that simply don’t apply.

You do have a company right now called Alembic that that just I mean, recently, like in the last few days, they debuted an AI that they are claiming eliminates hallucinations in enterprise data analysis. And right now, analysts and Fortune five hundred companies are kinda going nuts about it. And so that that’s kind of the big a big hurdle that a lot of people are really looking at solutions, and we’re seeing solutions happen all around us all the time. Challenges with the AI accuracy, absolutely a major drawback. And the good news is is that some methods are being introduced that are mitigating that dramatically.

So just so everybody I’ll get ahead of some questions that might pop up. You guys are gonna continually hear different names thrown out, and what you’re seeing is a whole new like, an expansion of an industry. I won’t call it a whole new industry. But if you start to look at the impact that AI is having, it took over the data center space in the last couple years.

Impact on Data Center Space

Every every AI it’s almost like back I compared to when Bitcoin came out, and everybody was looking for data center space to mine and do all the things. The biggest difference, though, these companies are well funded and backed by PE and other investment types, and they have, taken over. That’s why the data center boom has happened in the last couple years, and you’ve seen, all most of the space that’s allocated for twenty twenty four, twenty twenty five, already presold, and all of these these bills going on. So with that, you’re gonna see new companies names pop out.

You’re seeing a lot of a lot of, startups and everything in this space.

Emerging Suppliers and Channel Impact

We’re gonna find where they land. Some of them are gonna come directly into the channel. Some of them will be in the background and and be resources for some of our known suppliers to go and and grab the technology and implement it in and grab it as a service. So as this starts to emerge and you have questions about names, just know that we’re sourcing.

This is a major point of, emphasis for us as an organization to get you guys the technology in to sell to your customers. And And if we’re a little ahead of the game, we know how to prep the suppliers and go, hey. This is still an emerging space. The the bicycles are gonna be a little bit behind, but stick with us, get your name out there in the channel, etcetera.

So we’ve we’ve really nailed that down on on how we go to market with our with the emerging suppliers in the space. You’re also seeing suppliers that are well known to you develop AI products, especially in what Jason just laid out around the AI, making sure that they’re the the accuracy is there, creating pulling from data warehouses and data lakes and creating data lake houses that combine the structured and unstructured data for the ad to live in. And we can we can get into a deeper tech conversation than that and a and a one off training for anybody that wants to understand that landscape.

Data Concerns and Cost Issues

But it also is really nailing down the data concerns and cost issues. Now the cost issues is really gonna be you you’ve gotta start to tighten up value conversations versus cost conversations when it comes to AI because it is likely to be an investment this early on. There’s not a lot of organizations that probably budgeted for AI. But as they go into their budget cycles, that’s where you can start to get ahead of that for next year and things like that.

And what you can really start to work through is what the the bottom line effect is gonna have on this. And this is a this is a well known cloud approach as well when you start to sell on cloud. It’s the the force multiplier of the action. We’ll take, we’ll take Telarus.

I’ll do a one minute overview of Telarus.

We developed a lot of great tools early on. The way that we did it, infrastructurally, We weren’t able to grow and expand, so we had to go through a big transformation process over the last twelve to eighteen months. Our teams have worked really diligently. We’ve gotten to that place, which is gonna allow us to develop faster, create more products, get more things in your hand, and make sure that you guys have a better overall partner experience. That’s the kind of innovation that happens when you go through digital transformations.

Digital Transformation Impact

There was an investment and a cost, and then you can start to push through all the technology.

If you start to think of that think of it that way with AI and all the impact that it’s gonna have throughout, it’s not just gonna solve one problem most likely. It’s gonna solve a multitude of problems and have a greater impact. So even if there’s an investment, you gotta walk your customers through what their ultimate goal is gonna be. In fact, if you look at, if you go to the next slide, look at all the areas that it impacts.

You’re gonna see that it impacts a ton of different areas in the AI use and just in the cloud use cases. Then you throw in CX and cyber and so on. We’re just gonna call out one on the screen for for the sake of time, Fin ops, optimization of spending. That’s one of the biggest conversations that we have.

Use Cases and Impact Areas

There’s already AI that exists now that goes through and optimizes out, hey. You got these workloads going here. This would probably be better over here and things like that. That’s gonna get more and more refined.

The problem in the past from a perspective has been most of those organizations that developed that technology got bought by Microsoft or Amazon or or Google. And so they start to seem to be like, hey. Everything works better in our platform all of a sudden. Weird.

There’s some more agnostic ones and other ones that are continually being developed. If anyone’s ever had to sit down and look at an AWS bill, the AI that comes out that can just explain that in human is gonna be one of the will probably win some kind of a prize somewhere because it’s been one of the biggest frustration parts for every IT organization that utilizes AI as a platform. So you’re you’re gonna see a lot of use cases. It’s gonna be a lot to keep up with.

Take a breath. Remember, you have resources here that can walk you through it, and more importantly, help you walk your customers through it. It is gonna change your buy your sell motion a little bit too. Right, Jason?

That’s what that’s one of the big things that we’ve seen. If you go to the next slide, you start to see that the traditional buying motion, you know, is like this. So, Jason, I’ll let you kinda walk through the next couple of slides because I think you explained it better than I’ve ever heard it.

Yeah. Well, I mean, we’re used to people defining a problem and then going out and looking for a solution that fits the problem that they’re trying to solve. And so they’re trying to come up with the closest fit. You know, can I buy the thing that’s going to cover most of the problem that I’m that I’m trying to resolve here?

Traditional vs Customized Solutions

And then you’re taking a look afterwards at the result and figuring out if the need was filled, and then maybe you have to go back to b and try a different solution.

But then and if we could advance the next slide. If you start looking at AI, you’re having a little bit more of a custom, shall we say, development of tools. And it’s leading to defining the current state and then figuring out what the desired future state is. And And because AI is that piece in the middle and that you can fully customize what it is that it’s going to do and because it’s smart enough to be trained and and do all sorts of other advanced functionality, you end up then putting a solution in place that will lead you to the exact desired future state that you need. So the power in this is the definition of what the problem is, what the problem needs to turn into, and then you create the solution in the middle to make that work.

So it it is a different motion, and you’re gonna have a lot again, people are gonna try to get to the end state before they even actually really understand what they need it for because of the pressure and human nature. It’s going back to that FOMO type of type of approach. And at this point, AI is everywhere. You’re gonna continually hear it, and you might wanna have a key stat for you as another talk track. If you go to the last slide, Jason, you emphasize this in the oh, one more back, guys. Got a little overzealous there. There we go.

ROI of AI Investment

If you look at this, this is this one kinda blows everybody away. It’s from Microsoft studies. For every dollar invested in AI, it’s realizing an average return of three point five x. Right? So I will say, like, that is, to me, a great conversation starter of why companies see, if they’re not looking at it, might wanna start. I know we have a number of questions in the chat, and we have some other presenters on here that we need to get to. So, Jason, I I’ll hand it to you to kind of summarize or finalize any final thoughts on on AI as a service and the impact that it’s having.

Encouragement to Dive into AI

No. I would just encourage all of you as tech advisers to dive in, and don’t wait. Don’t be scared. We are here as Telarus resources to help you and to guide you and to do what we can.

We do have courses on Telarus University that can help you learn and familiarize yourself with some of the basics of artificial intelligence and how it actually works. And so go take advantage of that. And when you’re ready, reach out to your Telarus engineering group. We’ve got a lot of people that can help you out and help you get there.

Jason, Kobe, great job today. Fantastic presentation.

Response to Questions and Comments

We’ve used our time for q and a today, but we’re going to record the, questions that are in the chat window. We’re gonna respond to those individually later. Thank you all for your questions and comments on that, primarily around training, around, what’s hype and what’s real in AI and AI as a service.

But, obviously, a tremendous conversation that, partners can certainly be having with their clients. Kobe, any last words?

No. Doug, with Jason and I both being in Salt Lake, I think we can, grab some studio time, take these questions, and do a recording, and and go over it live and then push it out to the group. Might be the easiest way to do it. And then, come circle back with anybody that needs a a deeper dive. So thank you guys for having us.

Conclusion and Thank You

Wonderful presentation. Thanks, everybody.